Movement noise suppression in a moving array for downhole leakage localization
The disclosed technology provides ways to suppress or eliminate the effects of roadnoise when performing acoustic leak detection in a wellbore environment. In some aspects, a method of the technology includes steps for receiving acoustic training data, wherein the acoustic training data comprises signals representing acoustic tool contact with a wellbore surface, and generating a suppression model based on the acoustic training data, wherein the suppression model is configured to suppress roadnoise received at a hydrophone array disposed within the wellbore. Systems and machine-readable media are also provided.
1 . A method, comprising:
receiving acoustic training data, wherein the acoustic training data comprises signals representing extraneous noise signals;
generating a suppression model including an acoustic sensor signal correlation matrix based on the acoustic training data and machine-learning implementations, wherein the suppression model is configured to learn and model extraneous noise present in an environment of a wellbore, wherein the suppression model is based on a source correlation matrix, an interference correlation matrix, and a noise correlation matrix;
logging acoustic well log data from the wellbore using at least one acoustic sensor to acquire at least two measurements, wherein the acoustic well log data is captured at a depth corresponding to an acoustic well log operation of the at least one acoustic sensor;
adjusting one or more beamformer weights based on the suppression model for controlling suppression of extraneous noise included in the acoustic well log data;
processing the acoustic well log data based on the adjusted one or more beamformer weights to suppress the extraneous noise included in the acoustic well log data; and
processing the acoustic well log data based on the adjusted one or more beamformer weights to identify a leak or a flow in a casing within the wellbore or in proximity to the wellbore.
2 . The method of claim 1 , wherein the extraneous noise is generated during a down-logging operation.
3 . The method of claim 1 , wherein the extraneous noise is generated during an up-logging operation.
4 . The method of claim 1 , wherein the extraneous noise comprises roadnoise.
5 . The method of claim 1 , wherein the acoustic training data is generated in a laboratory setting.
6 . The method of claim 1 , wherein the acoustic training data is generated from real-world tool operations.
7 . The method of claim 1 , wherein the at least one acoustic sensor includes an acoustic sensor array.
8 . A system comprising:
one or more processors; and
at least one computer-readable storage medium having stored therein instructions which, when executed by the one or more processors, cause the one or more processors to:
receive acoustic training data, wherein the acoustic training data comprises signals representing extraneous noise signals;
generate a suppression model including an acoustic sensor signal correlation based on the acoustic training data and machine-learning implementations, wherein the suppression model is configured to learn and model extraneous noise present in an environment of a wellbore, wherein the suppression model is based on a source correlation matrix, an interference correlation matrix, and a noise correlation matrix;
log acoustic well log data from the wellbore using at least one acoustic sensor, wherein the acoustic well log data is captured at a depth corresponding to an acoustic well log operation of the at least one acoustic sensor;
adjust one or more beamformer weights based on the suppression model for controlling suppression of extraneous noise included in the acoustic well log data;
process the acoustic well log data based on the adjusted one or more beamformer weights to suppress the extraneous noise included in the acoustic well log data;
process the acoustic well log data based on the adjusted one or more beamformer weights to identify a leak or a flow in a casing within the wellbore or proximity to the wellbore.
9 . The system of claim 8 , wherein the extraneous noise is generated during a down-logging operation.
10 . The system of claim 8 , wherein the extraneous noise is generated during an up-logging operation.
11 . The system of claim 8 , wherein the extraneous noise comprises roadnoise.
12 . The system of claim 8 , wherein the acoustic training data is generated in a laboratory setting.
13 . The system of claim 8 , wherein the acoustic training data is generated from real-world tool operations.
14 . The system of claim 8 , wherein the at least one acoustic sensor includes an acoustic sensor array.
15 . A non-transitory computer-readable storage medium storing instructions for causing one or more processors to:
receive acoustic training data, wherein the acoustic training data comprises signals representing extraneous noise signals;
generate a suppression model including an acoustic sensor signal correlation matrix based on the acoustic training data and machine-learning implementations, wherein the suppression model is configured to learn and model extraneous noise present in an environment of a wellbore, wherein the suppression model is based on a source correlation matrix, an interference correlation matrix, and a noise correlation matrix;
log acoustic well log data from the wellbore using at least one acoustic sensor to acquire at least two measurements, wherein the acoustic well log data is captured at a depth corresponding to an acoustic well log operation of the at least one acoustic sensor;
adjust one or more beamformer weights based on the suppression model for controlling suppression of extraneous noise included in the acoustic well log data;
process the acoustic well log data based on the adjusted one or more beamformer weights to suppress the extraneous noise included in the acoustic well log data; and
process the acoustic well log data based on the adjusted one or more beamformer weights to identify a leak or a flow in a casing within the wellbore or in proximity to the wellbore.
16 . The non-transitory computer-readable storage medium of claim 15 , wherein the extraneous noise is generated during a down-logging operation.
17 . The non-transitory computer-readable storage medium of claim 15 , wherein the extraneous noise is generated during an up-logging operation.
18 . The non-transitory computer-readable storage medium of claim 15 , wherein the extraneous noise comprises roadnoise.
19 . The non-transitory computer-readable storage medium of claim 15 , wherein the acoustic training data is generated either or both in a laboratory setting and from real-world tool operations.
20 . The non-transitory computer-readable storage medium of claim 15 , wherein the at least one acoustic sensor includes an acoustic sensor array.